Font Size: a A A

Application Of SMCG Method And Preconditioning Technology For Microwave Integrated Circuits

Posted on:2006-03-04Degree:MasterType:Thesis
Country:ChinaCandidate:X P FengFull Text:PDF
GTID:2168360155958889Subject:Electromagnetic field and microwave technology
Abstract/Summary:PDF Full Text Request
This paper illustrates the analysis of microstrip structures using mixed-potential integral equation (MPIE) formulation solved with the Method of Moments (MoM). In recent years, analysis and design of planar structures such as microwave monolithic integrated circuits (MMIC) became a very important topic. Full-wave electromagnetic analysis of microstrip structures is especially significant. Due to the complexity of this scheme, efficient numerical tools are required for their analysis. The triangular discretization is employed to analysis the microstrip structures of arbitrary shape. One of the most popular techniques for analyzing this scheme is the sparse-matrix/canonical grid (SMCG) method. In this paper, we further improve computational efficiency of SMCG method by a preconditioning scheme based on the dominant information contained in the near field. We describe the near-field preconditioned sparse-matrix/canonical grid (PSMCG) that is effective in accelerating convergence of matrix equations for planar microstrip structures. Some hybrid techniques are developed based on sparse matrix iterative methods and sparse matrix direct methods, including conjugate-gradient (CG) method, symmetric successive overrelaxation (SSOR) preconditioned CG algorithm and multifrontal method. Our numerical calculations show that the PSMCG algorithms converge much faster than the SMCG algorithms for microwave circuits. Some typical microstrip discontinuities are analyzed and the good results demonstrate the validity of all these proposed algorithms.
Keywords/Search Tags:sparse matrix/canonical grid method, microwave circuits, near-field preconditioning, flexible generalized minimal residual, SSOR preconditioned CG method, multifrontal method
PDF Full Text Request
Related items